Evaluation of the WRF Model to Simulate a High-Intensity Rainfall Event over Kampala, Uganda

This paper aimed at evaluating the high-intensity rainfall data produced using the numerical weather prediction model, the WRF model, to be used for localized flood modelling in the data-scarce area. The performance of the model is evaluated by comparing different parametrizarion schemes as the combinations of microphysics, cumulus and PBL for sensitivity analysis as well as model simulation with and without considering the cumulus parametrization in the innermost domain of the model. The nature of this project's data is primarily numerical weather prediction model output with netcdf format. For the purpose of spatial model result evaluation, the open source daily satellite rainfall product called CHIRPS was used. Grid-wise model evaluation was used the observed rainfall data collected from a single gauging station in the study area. In the result, the best performing combinations out of the 24 different combinations were selected through TOPSIS statistical analysis For this particular event, the consideration of cumulus parametrization in the innermost domain of the model has also found to be less effect on model results.